CELL DIVISION IN EGGS OF CREPIDULA. 551 



of amitosis but a careful study of these nuclei which are connected by a chromatin 

 thread shows that in Crepidula at least, they are invariably formed in the manner 

 stated (figs. 141, 143, 144, 152-154, 215, 222). It is highly probable that the 

 number of chromosomes which go into two such daughter nuclei is not the same. 

 Such eggs do not give rise to normal embryos or larvae though subsequent cleav- 

 ages may go on in a manner which is approximately typical. (2) Apparent 

 amitosis due to the constriction of the achromatin and of the achromatic vesicle 

 following the mitotic division of the chromatin. It is evident that this is merely 

 a variation of the condition found in certain protozoa in which the nuclear mem- 

 brane does not disappear, and an intranuclear spindle is formed. In Crepidula, 

 however, the membrane does disappear, but if the daughter chromosomes are 

 prevented from absorbing achromatin, an achromatic membrane may form around 

 the achromatin of the spindle area, and in the later stages of mitosis such figures 

 may look like amitotically dividing nuclei (figs. 178-182). (3) Apparent amitosis 

 really due to the fusion of karyomeres. Many such cases are shown in my figures. 

 Karyomeres are formed whenever daughter chromosomes are slightly separated 

 into groups; this is especially the case when polyasters are present. These karyo- 

 meres may be very numerous, but those which are in contact fuse together dur- 

 ing the resting period, so that they become progressively larger and less numerous 

 in later stages of the division cycle (figs. 121, 124, 136, 145-150, 175, 177, 

 192-223). 



Much interest has been shown of late in the question whether amitosis occurs 

 in germinal or embryonic cells, since if it does it thereby weakens if it does not 

 destroy the belief in the chromosomes as the sole "bearers of heredity." I hold 

 no brief for this doctrine and have repeatedly urged (1893, 1899, 1905, 1908, etc.) 

 its too narrow outlook on the activities of the cell as a whole. It seems to me 

 incredible that this most general of all cell functions, which includes differ- 

 entiation, metabolism and reproduction should be the property of only a single 

 cell constituent, — the chromosomes. I have therefore approached this problem 

 without any preconceived bias as to the theoretical necessity of believing in 

 mitosis in all divisions of generative or embryonic cells, and have not been 

 consciously warped by either the odium mitoticum or the odium amitoticum. 



An excellent review of this whole subject is given in Wilson's book, The Cell 

 (1900), and in Godlewski's (1909) notable memoir on the inheritance problem, 

 and I shall confine my attention to a consideration of some of the more important 

 recent contributions on amitosis. 



After the discoveries which showed that mitosis was the usual form of nuclear 

 division, most of the workers who studied amitosis found it limited to cells 

 already fully differentiated and usually decadent. In recent years a new interest 

 has been given to the problem by the work of Child (1907 1 , 1907 2 , 1910, etc.), 

 Hargitt (1900, 1904), Patterson (1908), Glaser (1908), Gurwitsch (1905), Geras- 

 simoff (1892), Nathansohn (1900), Foot and Strobell (1911), etal, all of whom main- 

 tain that amitosis may occur as a normal process in germinal and embryonic cells. 



